On Fri, Sep 09, 2022 at 08:37:17AM +0200, Mattias Forsblad wrote:
quoted
quoted
+ chip->rmu.rmu_ops->get_rmon = mv88e6xxx_rmu_stats_get;
+
+ if (chip->info->ops->rmu_disable)
+ return chip->info->ops->rmu_disable(chip);
Why is a setup function calling disable?
So Vladimir Oltean commented before:
"I think it's very important for the RMU to still start as disabled.
You enable it dynamically when the master goes up."
This, plus the fact that mv88e6xxx_rmu_setup() already exists in the
tree, and calls chip->info->ops->rmu_disable(). It seems like that
doesn't need to change. Mattias is moving it around, and makes it seem
as if something is being changed. Maybe simple code movement could be
split into a separate change.